CHARLIE BENNETT: I'm going to turn off my mic so I can get completely clean audio. [MUSIC PLAYING] Stephen Hawking is still alive in data. He was using technology so much to create content, especially towards the end of his career. His words live on and will for infinity. [MUSIC PLAYING] (SINGING) I knew it must have been-- CHARLIE BENNETT: You are listening to WREK Atlanta, and this is Lost in the Stacks, the research library rock and roll radio show. I'm Charlie Bennett in the virtual studio with, going clockwise, Fred Rascoe, Ameet Doshi, Amanda Pellerin, and Marlee Givens. Each week on Lost in the Stacks, we pick a theme, and then use it to create a mix of music and library talk. Whichever you are here for, we hope you dig it. AMEET DOSHI: That's right, Charlie. Today's show is called "Metadata of the Dead Two-- Frankendata." [THUNDER] So Happy Halloween, everybody. Let's have a spooky show to celebrate. MARLEE GIVENS: "Metadata of the Dead Part One" concentrated on the social media that lives on after the users have died. What is Frankendata? FRED RASCOE: Well, if we compare Frankendata to Frankenstein, to make sense of it, I guess we're thinking of dead data that's been brought back to life. AMANDA PELLERIN: But let's remember, if we're following that plot line, then Frankendata data is actually the doctor. CHARLIE BENNETT: Oh, there's always somebody. [LAUGHTER] MARLEE GIVENS: Yeah, but you know, Frankenstein's data doesn't have the same je ne sais quoi as Frankendata. CHARLIE BENNETT: Let's think of it more like the data of the dead stitched together to continue as a monstrous accumulation. AMANDA PELLERIN: And we at Lost in the Stacks are chasing it with pitchforks and torches. AMEET DOSHI: Our songs today are about permanence, disappearance, and what remains when we're gone. MARLEE GIVENS: Oh, we sound spooky. Are we going to start with something creepy and eerie? AMEET DOSHI: I think we're going to start with something aggressive and loud, so that even when it's over and done, the sound will still live on as a ringing in your ears. CHARLIE BENNETT: Oh, like every concert I've ever been to. This is great. AMEET DOSHI: This is "Ace of Spades" by Motorhead right here on Lost in the Stacks. [MUSIC - MOTORHEAD, "ACE OF SPADES"] FRED RASCOE: Today's show is called "Metadata of the Dead Two-- Frankendata." You may remember the first "Metadata of the Dead" show from last Halloween, when we talked about what happens to social media accounts after a user dies. AMANDA PELLERIN: Our guest from that show has returned to continue the conversation. Robin Fay, a.k.a. GeorgiaWebGurl on the internet, has experience with libraries and the corporate world, and her mission is to help people use technology to make their lives better. She spoke with Charlie about how even after we die, personal data that we created and accumulated continues to live. CHARLIE BENNETT: So when we say personal data, what are we talking about? Because I can imagine quite a lot about when we'd say personal data-- maybe stuff that's embarrassing, things I've got in my notebooks up on the shelf over here. But when you think of personal data, what are you thinking of? ROBIN FAY: Well, I think of all the data that is accumulated by us, and then the data that we create. So by accumulated by us, that would be when someone tags us on social media. Or it could be if we have written something or published something or participated in a committee or organization, and our name has been associated with that committee or organization. Certainly, other data that's accumulated and created by others-- tax information that we get sent and gets compiled together, our medical information that gets compiled together-- all of that kind of stuff. And then, of course, things that we create ourselves. So our social media, stuff like that. CHARLIE BENNETT: And this is all kind of a loose conglomeration, right? This is not like some file somewhere that-- or folder that things are getting dumped in. Personal data is everywhere. ROBIN FAY: It is everywhere. And I think that is the challenge to really controlling your data. A lot of the data that we create ends up being part of public records. So if you buy property, or you maybe serve on some kind of committee, or you work in a government job where your information is part of the Open Records, or under the Open Records Act, then even your emails that you send, regardless whether or not they are work emails or personal emails, if you are using that computer-- you're using your work computer-- then sometimes those emails may fall into that category of being under the Open Record. So different people have different levels in terms of what kind of personal data is collected about them. But we all do have personal data collected by us-- by us and for us. CHARLIE BENNETT: Do you Google yourself to see if there's traces of you sort of scattered around? ROBIN FAY: Yes. And one of the things that I found sort of disturbing, and I know it's one of those things that a lot of people are concerned about, is things like your address being out on the internet. And in the olden days when we had telephone books, you could say I want to be unpublished, and don't put my address out there. But now, that information is considered sort of public records. And there's all these different companies out there-- if you just search someone it'll say, find out more about this person, or do a background check on this person. It's searching all the public records available. But it's also able to get other information that it purchases, things like if you have a landline, things like your phone number, your address, things like that. So it's amazing how much information is really collected by these various organizations, and then repackaged sometimes and sold without us having any control over that at all. CHARLIE BENNETT: Yeah. And so my next question was going to be, how does our personal data persist after death? But just listening to the description of personal data, it persists because it's disconnected from us, right? No one is waiting to hear your death notice so they can then erase what they've collected about you. It's all aggregated and floating around out there. ROBIN FAY: It is. And if-- you know, depending upon what websites, your information may appear on, they could even be archived at like the Internet Archive. They could be archived on a university website or a governmental website, something like that. So that information may persist really for all of eternity. And then, of course, we just had the census-- finished counting the census. Those census records get published online eventually. And so if you live long enough, you will actually see your name and information as a child appear on the census. So there's all this information that is collected about us, or that we provide to people, that then ends up on the internet. And yeah, there's no one who takes that information down. And I think where we notice it the most is on social media. So I was just looking through my Twitter list the other day. And I was like, oh, wow, there's one of my very early Twitter followers who has been dead now for 10 years. And I went back and looked. And it's like, wow, his wife posted in 2011 and kind of reminded people. But he had set up auto publishing. So he had been publishing-- I know I've talked about him before. But he's been publishing sort of-- CHARLIE BENNETT: Yeah. We've heard about him on our show. Yeah. ROBIN FAY: Yeah. He'd been sort of publishing from beyond the grave for at least a year before his wife kind of came in and shut everything down, and kind of explained to everybody what happened. And it was a very eerie experience. But I still see him in my Twitter feed. And then when I was looking through my Facebook, I noticed that there were actually two of my friends who have died in the last couple of years. And their pages are just there. They haven't been turned into memorial pages or anything like that. So that's the other part that happens is that it does create an extra-- almost a burden on whoever's left-- the friends or family or whoever's left of that person-- to sort of clean up their social media or let people know. Because now we have these networks of people who may not know that we're in the hospital. They may not know that there's been a death. They may have no idea of that information because the only connection they really have to that person is social media. So I think it really speaks to not only what the notion of friend is, but how that data is collected, and how that data is shared with people, too. There are some companies now that you can work with that will kind of work with you on your behalf to shut all these social media sites down. But there's really no one place where you can say to Google or anybody else, OK, please close out all of these social media accounts. Or say-- tell this person's followers and friends, we're sorry. This person has died and that kind of thing. So it is very interesting what the ramifications of someone dying who has so much information on the internet. And, of course, the other part of it then is it is nice to have an archive of that person, and to be able to have that information available and things like that. So I think it's one of the things that probably people don't think about too much. What are their wishes for their social media after they're gone? What are their wishes for the internet? [MUSIC PLAYING] AMEET DOSHI: This is Lost in the Stacks, and we'll be back with more about metadata of the dead-- metadata-- after our music set. FRED RASCOE: File this set under BQ4261 .R36. [MUSIC - MISSION OF BURMA, "SECRETS"] [MUSIC - ROBYN HITCHCOCK, "THE GHOST IN YOU"] (SINGING) --fade. Inside you the time moves, and she don't fade. The ghost in you-- MARLEE GIVENS: You just heard "The Ghost in You," by Robyn Hitchcock, and before that, "Secrets" by Mission of Burma. These are songs about things that seem permanent, but then disappear in a blink of the eye. [MUSIC PLAYING] FRED RASCOE: This is Lost in the Stacks. And our guest today is the metadata and technology librarian, Robin Fay, also known as GeorgiaWebGurl. She and Charlie spoke about data that lives after someone has died. CHARLIE BENNETT: So I had an experience recently of an older relative passing away. And this was a person who did not really interact with social media. I don't even know if there was a Facebook page. And I googled him just to see what was left over. And there were some entries in directories, like professional directories, that had gone from paper-bound to online a long time ago. So even the entries were kind of out of date. The most live stuff that I found of this person were comments on Google-- reviews of a couple doctors and of a car dealership. And beyond that, just nothing much. But I also have at home, personally, privately, a ton of pictures, a lot of emails, letters and all that kind of thing. And so I was struck at how it sort of-- getting ready to have this conversation with you, I was struck at how this profile of this person was what I think we expect to happen. Mostly, it's stuff that's at home behind the fire-- the physical fire wall of the private residence. But then comparing his online profile to even mine-- I'm mostly off social media-- but to my profile or other family members, there's so much stuff. If we just disappeared, it would continue for a while. Now, that feels like kind of a weird thing, an uncanny thing. But what is the actual problem with that? What is bad for society, I guess? Or what is dangerous about having this kind of Frankendata continuing even after someone has died? ROBIN FAY: Well-- and I think that some people would argue that there's nothing wrong with it, that we can live on. We can be infinite. And we can live forever on the internet. [LAUGHS] I mean, I think there are definitely people who would make that argument. And I don't know that there's anything necessarily wrong with it. But I do think there-- For example, if you have a Twitter account that is continuing to post after the person has died, I do think there is-- unless there's some way-- some kind of notification, I think there is a little bit of a misrepresentation, and it's just not an authentic experience at that point because you've lost that ability to be able to talk to that person and connect with that person. And then not knowing-- it's almost like talking to a robot or some kind of AI or something, in the fact that you don't really know who you're talking about and I-- talking to. And I would guess you could say that for a lot of people on the internet anyway. If you just email somebody or message them on the internet, you don't really know them anyway. But I don't think there's anything necessarily wrong with that. I just think that people need to have that-- people should have that information. CHARLIE BENNETT: And you're a fairly online person, too. I mean, you know the internet. You know how the technology works. You use data professionally. Do you think this is universal? Or is it particular to your interest in metadata that's making you really think about the inauthenticity of leftover-- of residual data? ROBIN FAY: Residual life. [LAUGHS] You know, I think-- I know when I've seen other people respond or comment when someone has died, and especially if it's something where now so many of the social media sites, when there's a lot of comments, then it will kind of float up to the top of everyone's feed, so everybody kind of sees it at that point. So, of course, when the statements made that this person has died, there's just a few comments. But then as more comments get added, then kind of everybody eventually sees it. And so it could be-- for example, when one of my cousins died, it was a couple of months before I knew about it. And the reason why I knew about it was it floated up to the top of my feed at that point. And then I saw it even later, like six months later, that showed off my feed again because somebody had commented and said, oh, I didn't know she died. And we all were kind of exposed to that again. So I think it does-- in terms of potential harm to society, I think the way social media handles those posts, if it's not made into a memorial account of some kind or shut down, is it can both help or hinder the grieving process. Because it can help the process in that you have this sort of collective appreciation for this person. You get to share images and memories and all this kind of stuff. But it may come up-- like, for the longest time, until I finally was able to say don't show these images to me anymore, when my dog died, it would be like, oh, once a month they kept trying to show me puppy pictures of her as a puppy. And it's like, oh, I got a new dog now. And I don't really-- I mean, I loved that dog, too. But I don't really need that coming up in my feed, you know? And it was kind of a little bit bittersweet. And so I do think that is one of the things that is an interesting part of this is because, one, if people really never die on the internet, then there's sort of that-- you don't really know if you're interacting sometimes with a live person or not, or engaging with a live person or not. And, in some cases, that probably doesn't make a difference. I mean, obviously, if you're dealing with some reporter or something like that, it's not typically in a friendship kind of aspect. So you're not going to have that kind of relationship necessarily. But then I do think when you start looking at how social media does push posts, especially memory posts, out to people, then it can-- I think it can cause that grief cycle to just kind of keep happening over and over and over again. Because it's like, oh, here's this person who just found out six months later, and now we're all talking about it again. And so I kind of-- it's going to be interesting to see, as we continue to grow with social media and more people die, what this means. CHARLIE BENNETT: Now, we've talked a lot about social media. What about that official data that you were talking about, or the interaction data, the people's internet browsing history that's on their ISP? Do you consider that part of this Frankendata? Do you think that gets used? Do you know what happens to that kind of floating data after you're dead? ROBIN FAY: It just seems like to me, from when I've sort of looked at the data as a whole, and I've looked at my data and all the different kind of data that's out there, it just seems like you can shut down specific-- or kind of say, like, hey, this is done, or make this a memorial page or whatever. But there's so much data out there that you can't do that with. And then when you start looking at things like author identities, like an ORCID, or you look at author identities in ResearchGate, or academia.edu, or the Library of Congress National Authority File, or all those sort of official places that we put identity information in, it then really relies on information professionals-- librarians, metadata workers, people like that-- to say, this person has died, and we need to change this information. Wikipedia-- the same way. Wikipedia is a little bit more general. Kind of anybody can make a recommendation, make that edit. But still, somebody has to go in there and make that edit and say, this person's died. In the case of Wikipedia, they actually have to write something about that. This person died. And they usually write something about the circumstances. That's a manual process that involves people. And it's true when it comes to things like Wikipedia, it's true when it comes to ORCID, or the Library of Congress Name Authority Files, it relies on people. AMANDA PELLERIN: We'll be back with more about social media and death on the left side of the hour. Dark. [MUSIC PLAYING] FREDERICK ARNOLD: Hey, guys. This is Frederick from Future Oak Records. [MUSIC PLAYING] You are listening to Lost in the Stacks on WREK Atlanta. [MUSIC PLAYING] (SINGING) Dry the sheets. Place the books. Sit tight. CHARLIE BENNETT: Today's Lost in the Stacks is called "Metadata of the Dead Two-- Frankendata." And I'd like to read a bit of an article titled, "What Really Happens to Your Big Data After You Die?" by Bernard Marr published in Forbes in February of 2017. Marr is describing strict rules regarding medical data. And then he switches gears to say, "This tight regulatory framework, however, certainly isn't in place when it comes to all the personal data which is collected by private companies and organizations. Google is an obvious case. Unlike medical data, there's far less regulation around what is done with the sort of data they collect. Although emails, GPS, documents, and financial information can often be just as personal and private, Google puts no limits on how long it will hang on to this information after someone dies. On top of this, there's not even an obvious mechanism to let them know that the data they have belongs to a dead person. So they will presumably continue to handle it in exactly the same way they did when its owner was alive." And you, dear listener, can add the editorial comment to that that you wish. File this set under BD444 .M3. [MUSIC PLAYING] (SINGING) for good. AMEET DOSHI: That was "Gone for Good" by Poison Idea. Before that, "Shed the World," by Lungfish. Those were songs about leaving and what's left behind. [MUSIC PLAYING] Our show today is called "Metadata of the Dead Two-- Frakendata." And our guest is Robin Fay, a.k.a. GeorgiaWebGurl on the internet. CHARLIE BENNETT: Now, I feel this is the time for me to ask the question, what does this have to do with controlled vocabularies? And what are controlled vocabularies? ROBIN FAY: Well, on the simplest level, controlled vocabularies is just a list of topics that someone would pick off. So, like, a list of topics-- if you're doing blog posts you might choose some kind of category or sort of topic to describe that particular blog post. In terms of controlled vocabularies on how this would work, really looks at people and what their identities are. So this is one thing that the internet has not done well as a whole, in terms of establishing individual identities. We see it a little bit with Wikipedia. And the Wikidata project is starting to really work on that. But if you go into Wikipedia and look you'll see, for example, James Brown. You'll have James Brown the musician, and James Brown somebody else. So it tries to separate out all of these identities. Because if you just Google James Brown, you're going to get a whole bunch of different people in Google. Libraries have done this for a long time. We say James Brown the musician. And we create an identity record, and manage that identity to explain exactly, this is the James Brown-- this is the musician versus the artist. The internet hasn't typically done that very well. So that's how the controlled aspect works of it, is knowing that this person that exists-- what their identity is. And then part of that being tied to that identity is things like the birth and death dates. CHARLIE BENNETT: Yeah. I was thinking about how some version of Frankendata is actually the texts that are contained in a library, and how they have to be identified, and the status of the author needs to be recorded. So when you see a subject heading or an author record that has the date of birth and date of death in there, it's sort of like the library acknowledges this is data continuing after someone has died. But we're going to let you know that they died. And this is the stuff that continued on purpose. And it seems like we have a lot more of nonpurposeful data continuing after death. ROBIN FAY: Yeah. And I think that the key part of it for library data in terms of this is we really use the date as an identifier. So that is a way to kind of shape the identity of that particular person. So when we are looking at how do you determine James Brown, the musician? Well, I mean, obviously, he's the musician. But then what other information can you use to kind of explain that identity? And birthdate is one of the easiest ones typically to get-- not always. But between census and other information, artist's pages, bio pages, things like that. Publishers provide that information for authors a lot of times. That birth date or death date is a really easy thing to collect. So that piece of it is so important in terms of that identity management part of it. And, like I said, on the larger web, we're starting to look at that. If you go in and you search for people-- and I'm one of the people if you search for me, a Google card now comes up, and it gives some information about me and-- some very basic information. But that has not been the way the internet was kind of created from the beginning. The internet was very decentralized. Data was very decentralized. And we do seem to be moving to more structured data as we go along. And there have been several attempts in the internet to do that, to sort of make data more structured. CHARLIE BENNETT: Do you like that? As a personal internet user and as a professional, do you like that more structured internet? ROBIN FAY: The advantage to having a more structured internet is that searching becomes more accurate, and you have a lot more ability to search more granularly. So you can really find what you need. Now, there's a lot more that goes into searching with Google. There's paid ads and all that kind of stuff that goes on. But just in theory, more structured data should lead to more relevant results, in theory-- taking out all the other pieces of it. And I think that's one place definitely where library data has really excelled is that we do have very structured data. We have issues with our data format. Doesn't play well with others. It's hard for search engines to use it, all that kind of stuff. But the structured part of the data and the kind of data that we use, it's just very robust compared to what is out sort of in the general web world. CHARLIE BENNETT: So if information professionals are sort of moving towards this structure, like always trying to decrease entropy in the internet data, what can we do, i.e., what can people who are just internet users, maybe not information professionals, what can they do to mitigate Frankendata, or to help keep useless or purposeful data from continuing on after they are gone? ROBIN FAY: Well, I think there are so many different things that people can do. In terms of unuseful data, that can even be things like fake news. I mean, it can go much beyond just sort of personal data, when you're sort of looking at that data that is not useful. Certainly, almost all social media platforms offer a way to report things. If you are a social media user, you can actually set up a person to serve as sort of your contact person, or to oversee your account in case something happens. So you can set up someone to take charge of that. I know people are starting to write it into their wills, in terms of including that information-- what's my social media plan? In terms of when you see data on the web that's outdated, if there's a way to be able to report that. And you even report things at the Library of Congress website. If you're searching their library catalog, there's even a place to report something there. So I think there are reporting mechanisms. And I think that's useful to do for libraries. We have done a very good-- or typically, do a very good job of doing what some people would call close out a name or add the death date. When someone dies, there's, in fact, usually, sort of a rush among information professionals to update the records in the Library of Congress authority file. And I'm sure Wikidata is probably the same way, where it's like, oh, we need to get that name-- the date on there. Sort of a badge of honor. Whoever gets to add that death date first-- a weird badge of honor. But so I think there is that part of it. And then the other piece of it is as we get more structured data, and as we start to look more at identity management across the web as a whole-- libraries, like I said, have done a fairly good job of that. But as we look more at identity management across the web as a whole, then there is what can we do to streamline the data and keep the data in sync? And that's where this notion of linked data comes in, that you're linking to a source, rather than actually typing in that information. So for James Brown, the musician, you would link to the Wikipedia James Brown entry. Or you could link to the Library of Congress name authority file for James Brown, the musician, rather than typing in James Brown musician and his dates. CHARLIE BENNETT: So I have a last question for the show. It's more of a personal question. What have you put in place? What's going to happen when you die? [MUSIC PLAYING] ROBIN FAY: Well, in terms of all my social media, there is a plan in place. You know, I do occasionally have automated publishing from things. So that's something to think about. I used to use some actual tools to schedule things out for weeks and sometimes months ahead. So that is something that I'd still need to think about. But yeah. I feel like I have a plan for all of the social media aspect of it. I'll leave it up to you all [LAUGHS] to figure out the rest of the data. [LAUGHS] CHARLIE BENNETT: Well, Robin, thanks for being on the show today, and Happy Halloween. ROBIN FAY: Thank you. You too. MARLEE GIVENS: This is Lost in the Stacks. Our guest is the librarian, technologist, and trainer, Robin Fay, a.k.a. GeorgiaWebGurl, on the internet. We'll be back after a music set. AMEET DOSHI: File this set under BF1461 .R8. [MUSIC PLAYING] (SINGING) You've got your high-heeled sneakers on. [MUSIC PLAYING] (SINGING) Always something there to remind me. AMANDA PELLERIN: You just heard "There's Always Something There to Remind Me," by Sandy Shaw. And before that, "Out of Sight, by James Brown-- songs about what's hidden and what remains no matter how hard we try to remove it. [MUSIC PLAYING] CHARLIE BENNETT: Today's show is called "Metadata of the Dead Two-- Frankendata." So let's do a lightning round, everybody. [THUNDER] What's the kind of data that you create, without actively trying to create it, that will persist after you're gone that makes you feel weird or scared or spooked? Mine is financial transactions. I start to freak out when I think about how many little bits and bytes about what I've bought, or tried to buy, exist out there. AMEET DOSHI: For me it's the number of times I've gone to newyorktimes.com without reading it, and quickly closing the tab before seeing what the news of the day is. I think that's a frightening number of times this year. MARLEE GIVENS: That's so specific. I think mine is all the times that I have inadvertently clicked on click bait. AMANDA PELLERIN: Mine has to be my Amazon order history. Amazon called me one time asking if I wanted to open a business account, which was very startling because I didn't even know that Amazon had customer service people. [LAUGHS] Oh, no. I have a problem. FRED RASCOE: I'm having a hard time picking mine because all of it kind of makes me feel creeped out and a little disturbed. I don't know if you remember that movie Coco that came out a couple of years ago. But in that movie, people that die go to a spirit world and remain there as long as someone remembers them. I fear that in this digital environment, that spirit world is going to just become overpopulated, and there will be no blessed fall into oblivion. We'll always be there, and we'll always be remembered by our virtue of our digital data, or Frankendata, that lives on. CHARLIE BENNETT: Well, with that, let's roll the credits. I think I got some more Lemmy here. [MUSIC PLAYING] FRED RASCOE: Lost in the Stacks is a collaboration between WREK Atlanta and the Georgia Tech Library, written and produced by Ameet Doshi, Amanda Pellerin, Charlie Bennett, Fred Rascoe, Marlee Givens, and Wendy Hagenmaier. MARLEE GIVENS: Today's show was recorded in the virtual studio, edited and assembled by Charlie, and brought to you in part by the library collective and their social and professional network league of awesome librarians. Find out more about them at thelibrarycollective.org. FRED RASCOE: Legal counsel and a controlled vocabulary for grapes and soils provided by the Burrus Intellectual Property Law Group in Atlanta, Georgia. AMANDA PELLERIN: Special thanks to Robin for being on the show. And ironic thanks to all the folks working to keep our data separate from ourselves. And thanks for real, as always, to each and every one of you for listening. MARLEE GIVENS: Find us online at lostinthestacks.org. And you can subscribe to our podcast pretty much anywhere you get your audio fix. CHARLIE BENNETT: We have a rerun next week. And the week after that, we're talking once again about bodies on the line. AMEET DOSHI: It's time for our last song today. The Frankendata that remains after death is a heavy subject. It's scary to think about all that data about us and how it remains after we're gone. And to be honest, 2020 is the kind of year that's given me too much to think about already. So I'm going to reflect on some positive things now, like favorite candies from my youth-- Big League Chew, Now and Laters, Jolly Ranchers, and not think about the negative, like all of that stuff is congealing in my veins. This is "Don't Think About My Death," by The Wood Brothers right here on Lost in the Stacks. Pixy Stix. Have a great weekend, everyone. [MUSIC PLAYING]